Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 16 Nov 2014 17:45:56 -0800
From:      Steve Kargl <sgk@troutmask.apl.washington.edu>
To:        Hans Petter Selasky <hps@selasky.org>
Cc:        Adrian Chadd <adrian@freebsd.org>, freebsd-current <freebsd-current@freebsd.org>, Dag-Erling Sm?rgrav <des@freebsd.org>, Ian Lepore <ian@FreeBSD.org>
Subject:   Re: USB locks up system -- WAS Re: shutdown or acpi problem
Message-ID:  <20141117014556.GA38184@troutmask.apl.washington.edu>
In-Reply-To: <5468FBFB.1090208@selasky.org>
References:  <1416162070.4781.186.camel@revolution.hippie.lan> <F0B9570C-BADD-4843-AE91-66CAB0B5DE85@FreeBSD.org> <1416162718.4781.192.camel@revolution.hippie.lan> <4F3B4E02-0ADD-46EB-BB90-BF360E810585@FreeBSD.org> <20141116185141.GB36021@troutmask.apl.washington.edu> <4443FEFB-17F1-4516-959F-4C70AB2A4382@FreeBSD.org> <20141116190344.GC36021@troutmask.apl.washington.edu> <5468F814.6010702@selasky.org> <20141116192959.GB36339@troutmask.apl.washington.edu> <5468FBFB.1090208@selasky.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Nov 16, 2014 at 08:33:15PM +0100, Hans Petter Selasky wrote:
> 
> thread apply all bt
> 
> That will give you the backtrace of all threads. Grep for usbconfig, and 
> figure out which line is causing the problem in the kernel. Then look at 
> the USB explore threads and see where they are stuck in the detach of umass!
> 

I haven't tried the kgdb approach, yet.  I can say that the
problem is coming from sg device.  If I have 'device sg' in
my config file, the problems occur.  Removing 'device sg'
gives a working kernel.

If I do not start hald in /etc/rc.conf, everything works as 
expected.  Now, if hald was started at boot and I then stop
hald with '/usr/local/etc/rc.d/hald stop' then the one hald
relate process is not stopped. I've wrapped lines to keep this
short.

% ps axl | grep hald
 UID  PID PPID CPU PRI NI   VSZ   RSS MWCHAN   STAT TT  TIME      COMMAND
  0  1058    1 195  20  0 13180  3512      -      I  -  0:00.05
      hald-probe-storage: /dev/sg1 (hald-probe-storage)

Using the procstat as suggested by Adrian, I have

% procstat -ka 
1058 100157 hald-probe-storage hald-probe-stora mi_switch sleepq_switch
     sleepq_catch_signals sleepq_wait_sig _sleep sgread giant_read
     devfs_read_f dofileread kern_readv sys_read syscall Xint0x80_syscall 

-- 
Steve



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20141117014556.GA38184>